Conditional generative adversarial network for 3D rigid-body motion correction in MRI.
CONCLUSION: The images predicted by the conditional generative adversarial network have quantitatively and qualitatively improved image quality compared to the motion-corrupted images.
PMID: 31006909 [PubMed - as supplied by publisher]
Source: Magnetic Resonance in Medicine - Category: Radiology Authors: Johnson PM, Drangova M Tags: Magn Reson Med Source Type: research